How to recognize product based on image using neural network?

Question

Company has many products in their offer (some about 100,000), some of these are very similar to each other. In database there is available only one image per product.

Company want to make possible to recognize product based on video camera and display its specification. Is it possible to train new model with this kind of data or build it using existing model?

Answer 1

Is it possible to train new model with this kind of data

Yes, you need Convolutional Neural Networks (CNN) for image classification. If you only have one image per product in your dataset, I suggest you to use a lot of image augmentation, a technique that is meant to artificially increase the size of an image dataset by applying combinations of distortions to image data.

---

or build it using existing model?

You can re-train existing models. On the TensorFlow website you can find a huge list of pretrained CNN architectures that you can download ad use for your needs. It's a powerful alternative to training a new model from scratch.